摘要:

R. A. Bennett Surface Science and Catalysis Research Centre, Department of Chemistry, University of Reading, Reading, Berkshire, UK RG6 6AD. E-mail: r.a.bennett@reading.ac.uk Received 10th March 2000, Accepted 29th March 2000, Published 5th April 2000 Re-oxidation of reduced oxide surfaces is a fundamental step in catalysis and related fields. This paper reports the first study of the re-oxidation, at elevated temperature, of crystallographic shear planes (CSPs) occurring on the non-stoichiometric rutile TiO2(110) surface by scanning tunnelling microscopy (STM). The re-oxidation occurs by the reaction of interstitial Tin+ ions dissolved in the crystal bulk with oxygen at the surface to re-grow new TiO2 layers on the surface. The CSPs act as nucleation centres for the re-growth and are, in general, preserved during the early stages of the re-oxidation.CSPs appear in pairs on the surface and sandwich a layer of rutile that is displaced by one half an atomic step height from the original terrace. Each step edge of the pair shows markedly different structure and growth rate. The growth of the surface also takes place on the displaced region between CSP pairs. The re-oxidation of the substoichiometric TiO2(110) surface in the presence of crystallographic shear planes alternating (1 × 1) and (1 × 2) reconstructed surface structures at higher temperatures.17,18 In both cases the common cause is the re-growth of TiO2 on the surface due to recombination of Ti interstitials dissolved in the crystal lattice with O2 at the surface.At low temperature this leads to kinetically hindered, poorly ordered structures while higher temperature allows the formation of well ordered surface reconstructions. We have recently investigated the re-oxidation of the cross-linked (1 × 2) reconstructed surface in detail17 and will only highlight the key points here: (i) upon exposure to oxygen the missing rows of the (1 × 2) surface begin to fill in with an increase in the number of cross-linking structures; (ii) the filled in regions rapidly restructure to form (1 × 1) islands embedded in the original (1 × 2) reconstructed terrace; (iii) the (1 × 1) islands grow in size and merge to form (1 × 1) terraces upon which new TiO2 grows as bright points and rows; (iv) these bright features grow in length and number to form a new cross-linked (1 × 2) surface; (v) this cycle then repeats for the new layer; (vi) step edges show (1 × 1) termination at the upper edge and also move across the surface, predominantly by coalescence with the (1 × 1) structures on the lower terraces. 1.99.The accommodation of departure from stoichiometry by dissolution of Ti into the bulk can lead to dramatic structural changes as the stoichiometry approaches TiO Further reduction below this level leads to the formation of very dark blue/black crystals, which naturally show the cross-linked (1 × 2) reconstruction after annealing, with additional linear structures which are not aligned with the principle surface vectors.These lines are due to formation of crystallographic shear planes (CSPs) in the bulk of the crystal which terminate at the surface as a half height step. CSPs are a form of planar defect clustering common to reduced d0 oxides of Ti, V, Mo and W.1 For the case of TiO2, CSP formation can be visualised as a change from the normally edge sharing distorted octahedra (Ti at the centre, oxygen atoms at the apices) in stoichiometric rutile to a face sharing arrangement at the CS planes, see Introduction TiO2 is employed in a diverse range of applications, which has led to it becoming by far the best-characterised oxide surface. Titania is currently used industrially as a white pigment in paint and cosmetics, as a support for catalysts and photocatalysts, as a gas sensor and as a biocompatible interface for medical implants.These disparate functions are critically determined by the surface properties of the oxide.1 The very nature of the oxide surface dictates that the most active site for a chemical reaction is often at a surface defect and such defects may be present only in low concentration or even transiently. The recent advent of STM, and related probe microscopies, has allowed investigation of the chemistry of these surfaces at the atomic level and the role of minority defects to be assessed. The surface chemistry of the (110) face has received considerable attention1–21 and can be prepared in a number of surface terminations depending upon crystal preparation and history.17–19 The bulk truncated (1 × 1) termination is produced by annealing crystals with small departures from stoichiometry in O2 at elevated temperature.This procedure leads to the formation of blue crystals with low bulk defect concentrations. For more heavily reduced (dark blue) crystals both true and cross-linked (1 × 2) surface reconstructions evolve which have been related to added rows of stoichiometry Ti2O3 for the true (1 × 2)3,4,21 and Ti3O5 or TiO2 for the cross-linked (1 × 2).6,7,10,11,17,18 STM imaging has shown that both the structural components of the added rows straddle the bright rows imaged on the (1 × 1) surface. The bright rows of the (1 × 1) surface have been shown to be due to imaging of empty states of the 5-fold coordinated Ti4+ ions.8,9 Typical images of these surface structures and associated schematics are shown in Fig.1. The re-oxidation of the reduced crystals in a low pressure of oxygen is a common procedure carried out to prepare well ordered (1 × 1) surfaces. Recent STM experiments have shown that the re-oxidation itself may generate new surface structures (at low temperature)16,17,19,20 or form DOI: 10.1039/b001938k PhysChemComm, 2000,3Fig. 1 STM images (200 × 200 Å, 1 nA, 1 V) and schematic diagrams of the bulk terminated (1 × 1) (300 K) and reconstructed (1 × 2) surface structures (image C taken at 673 K). (A) The bulk terminated surface showing rows of 5-fold coordinated Ti4+ ions running in the <001> direction. The unit cell is marked in the schematic in plan view.The lower side view shows the bulk crystal structure with the red triangles highlighting the distorted octahedra centred on the Ti4+ ions. (B) Added rows of Ti2O3 running in the <001> direction that forms the true (1 × 2) reconstructed surface. A single row is shown in the schematic views. (C) The cross-linked (1 × 2) surface composed of added rows of TiO2 in their normal bulk positions. The rows are periodically cross-linked every 12 lattice units to form an ordered surface structure. The proposed structure of the cross-links is shown circled in the schematic. The horizontal line across all schematics is a guide to the eye to highlight the differing physical heights of the surface corrugation. Click image or here to see a larger version.ref. 1, 22–24 for detailed structure. The formation mechanism is believed to involve diffusion and aggregation of Tin+ interstitial ions to form a critical defect nucleus.24 This nucleus then drives a restructuring of the lattice that expands to form a pair of CSPs in which the rutile structure between the CSP pair is displaced by 1/2 <011> with respect to the bulk crystal. Half height steps are generated on the [110] surface at the CSP. There is no net displacement across a pair of CSPs, reducing the intrinsic strain associated with accommodating high concentrations of interstitials (some of which are bound within the CSP). Where the departure from stoichiometry is large the CSPs can order into arrays with regular spacing to form Magneli phases.25 The observation of the surface termination of the Magneli phases of TiO2 by STM and LEED has received some attention in the literature,26–29 however, there has been no detailed determination of the surface structure.Investigation of the surface chemistry of the CSP structures has received little attention despite the formation of CSPs in TiO2 supports of real catalysts,30 and the suggested role that CSPs play in partial oxidation reactions on MoO3 catalysts.31 In the latter example the reaction chemistry is thought to be dominated by the exposed surface sites that are produced when the CSP terminates at the surface. In this paper we present STM images and a time lapse movie of the re-oxidation of the titania at the CSPs and discuss the role of these special surface sites in relation to the chemistry of TiO2.Experimental The experiments were performed with an Oxford Instruments variable temperature STM additionally fitted with facilities for Ar+ ion sputtering, LEED/RFA Auger (VG Scientific), temperature programmed desorption and gas dosing. Tips were prepared by electrochemical etching of W(97%)Re(3%) alloy wire of 0.2 mm diameter. A more detailed description of the apparatus may be found elsewhere.32 The TiO2(110) single crystal (PI KEM, UK) was prepared by repeated sputter (600 eV, 1000 K) and anneal (1200 K) cycles to produce a dark blue/black crystal. During the initial preparation stages a (1 × 1) surface could be prepared by annealing at 1200 K, however, with time this procedure produced predominantlythe Ti2O3 type (1 × 2) termination.After even more cycles of sputter/anneal treatments only the cross-linked (1 × 2) surface could be prepared. Images are acquired at elevated temperature. There is no compensation for lateral thermal drift other than moving the tip slightly between images to maintain imaging of the same area. All images are taken with a positive sample bias; i.e. empty states of the surface are probed. Bias voltages given in the figure captions correspond to sample biases. Results and discussion Identification and surface structure of crystallographic shear planes The symmetry of the rutile crystal dictates that CSPs may form in four equivalent directions which complicates the identification of the CSP terminating at the [110] surface.Taking the general form of a CSP family to be composed {hkl} = p[121] + q[011] (p and q positive integers)22 the line of intersection of the CSP on the [110] surface is given by [hkl] × [110]. There are eight symmetrically equivalent planes in the {hkl} family ([hkl], [ l k h ], [ l k h ], [ l k h ], [khl], [ l h k ], [ l h k ] and [ l h k ]) giving rise to four lines of intersection from the cross product above, < x x x >, < y x x >, < xy x > and < xx x >, where x = p + q and y = 3p + q. From this we see that all values of p and q used to define a family of CSPs will generate a CSP with a symmetry which intersects the surface in the < x x x > and < xx x > directions.Thus the observation of CSPs in these directions at the surface does not allow a unique determination, as any family of CSPs may be present. However, observation of lines of the form < y x x > or < xy x > allows the simplest form of the CSP to be deduced as we can solve for p and q. An example of this is shown in Fig. 2 which shows three different types of CSP terminating at the surface of the (1 × 2) reconstructed surface. The uppermost CSP runs in the <111> direction and so cannot be identified whereas the remaining CSPs run in the <112> and <335> directions and so belong to the {132} and {143} family of CSPs respectively. Fig. 2 STM image (1000 × 1000 Å, 0.1 nA, 1 V, 773 K) showing the cross-linked (1 × 2) surface with diagonal crystallographic shear planes running across the surface.The direction of the CSPs is shown by <xyz>, while the simplest family to which the CSP may belong by {hkl} (where known). The blue rosette marks the same place as in Fig. 5. One further detail that needs to be noted is the nature of the interface between the CSP and the surface. As previously mentioned the CSPs form in pairs and represent the boundary between the normal rutile lattice of the bulk and the displaced (by 1/2 <011>) rutile within the CSP pair. Due to the shear planes running at an angle to the surface plane the two CSP 1/2 height step edges are inequivalent.This is demonstrated in Fig. 3a, which shows an STM image of a pair of CPSs running in the <335> direction with the upper CSP 1/2 height step edge having good epitaxy between the displaced (1 × 1) central slab and the (1 × 2) reconstructed terrace. In contrast at the lower CSP the (1 × 1) slab is poorly connected to the neighbouring (1 × 2) terrace with only a few points of contact. The surface structure in this region is complex; there is a down step from the central slab to a thin section of (1 × 1) terrace (black in the image). This (1 × 1) surface is the original layer upon which the (1 × 2) reconstruction forms. The reconstructed (1 × 2) rows preferentially terminate on this terrace rather than extend to meet the lower edge of the CSP slab.The schematic of Fig. 3b, shown side on to the surface looking down the <001> direction, shows how such differing structures result from the termination of the planes with the surface. On the left hand side there is a depiction of a normal crystallographic step edge which would be imaged as ~3.2 Å displacement by STM. The dark grey sections of crystal show how a CSP may terminate at the surface (for clarity the interlocking lattices of the bulk crystal and CSP slab have not been displayed as reconstructed). Furthermore there is no net displacement in traversing both CSPs, the lattices at the left and right extremities are in phase. At the surface both CSP induced step edges have 1/2 the normal crystallographic step height.However, the steps themselves have markedly different character—the left hand step moves up from a horizontal to a vertical octahedron while the right hand step down is composed of two vertical octahedra. It is therefore likely that differing surface terminations, reconstructions and facets may evolve due to the different free energies of the step edges. Most CSP pairs observed in this study show some degree of difference between upper and lower CSP step edges. These sites also have fundamental importance as the breaking of the symmetry of the surface has implications for reactivity, selectivity and especially for enantioselectivity in the case of chiral reactions. For the high bulk densities of CSPs formed in Magneli phases the reconstructions and step interaction of the CSP step edges are complex and lead to facetting of the surface of the crystal.29 Re-oxidation in the presence of crystallographic shear planes Fig. 3a showed the cross-linked (1 × 2) reconstructed surface exhibiting {143} and {132} CSPs maintained at 773 K.Movie 1 now shows high resolution images of the same area of surface during reaction of the surface with an oxygen overpressure maintained at ~2 × 10–7 mbar. Each image takes around one minute to acquire and save so the images form a time lapse movie. The initial frames (1–22) show the growth of the (1 × 2) surface up to the edge of the uppermost CSP pair and the formation of bright features on the CSPs. These bright features nucleate the growth of (1 × 1) islands within the (1 × 2) terrace adjacent to the CSP.The larger (1 × 2) terraces show an increase in the number of cross-linking structures and the normal step edge in theFig. 3 (a) STM image (400 × 400 Å, 0.2 nA, 1 V, 773 K) showing a CSP pair running in the <335> direction across the reconstructed terraces. Note the differing level of attachment of the (1 × 2) terrace to the raised central slab of (1 × 1) surface between the CSPs. Also visible is a change of the cross-links of the terraces from a full cross to a single link in the proximity of the CSPs. This may be related to long range strain fields produced by the CSPs. (b) Schematic showing a <001> projected view of a pair of CSPs terminating at the [110] surface.Oxygen ion distorted octahedra centred on the Ti4+ ions (dark dots) are indicated by the diamond shapes (cf. Fig. 1). The interstitial Ti populates the <001> channels in the bulk lattice as indicated in the bottom left while a normal step edge is shown top left. The distorted octahedra forming the CSPs are shown in grey and have not been relaxed from their bulk positions in the normal and displaced lattices. Where the CSPs terminate at the surface 1/2 height steps are formed which have differing structures. upper left corner begins to grow through the (1 × 2) terrace. The (1 × 1) islands grow along the CSPs with a bright line delineating the upper CSP pair. The (1 × 1) island notably grows preferentially from just one edge of this CSP pair.AVI motion picture (5 Mb) GIF motion picture (4 Mb) During oxidation (frames 22–32) the growth of the surrounding terrace forces the lower CSP pair to form a trench in which the CSP central slab is a 1/2 step down with respect to the (1 × 1) terrace. A new layer begins to grow on top of the CSP central slab by linking the neighbouring terraces by a short ridge (frames 34–46). The ridge is 1/2 step above the neighbouring terraces and with exposure extends laterally along the CSP pairs to form half height raised step edges (frame 47 onwards). Growth can then take place from this new raised ridge in the form of (1 × 2) strings extending from the CSP across the terraces in the <001> direction (frame 70 onwards). The growth of a new layer of TiO2 on top of the (1 × 1) surface is nucleated at the upper edge of the CSP pairs and from bright points on the newly formed (1 × 1) terraces (frame 60 onwards).These features extend to form bright strings which, if left to grow fully, would make up a new (1 × 2) surface. The final result is the production of (1 × 1) ridges along the CSPs, bright strings and (1 × 2) islands growing from the upper 1/2 height step edges of the CSP pairs. Movie 1 STM movie showing the reactive re-oxidation of crosslinked (1 × 2) reconstructed TiO2 in the presence of CSPs. All images are 300 × 300 Å taken at 0.1 nA and 1 V while the sample was maintained at 773 K in the presence of 2 × 10–7 mbar O2. Click the links above to access the movie.Fig. 4 STM image (200 × 200 Å, 0.5 nA, 1 V, 773 K) showing the state of the surface and CSPs after re-oxidation.The CSPs have been maintained on the surface and are now visible forming a ridge of (1 × 1) which has (1 × 2) strings attached to one (upper) boundary, while the lower edge shows a differing termination. The two different shades of (1 × 2) strings result from the presence of both Ti2O3 and TiO2 added surface structures (cf. Fig. 1). Fig. 5 Large area STM image (1000 × 1000 Å, 0.1 nA, 1 V, 773 K) of the surface after reaction. The blue rosette marks the same position as in the initial image prior to reaction, Fig. 2. The surface has undergone much restructuring with the movement of step edges, transformation of (1 × 2) to (1 × 1) and the beginnings of growth of a new layer.The CSPs in contrast have not moved position and are showing differing reconstructions on upper and lower CSPs of the pair. The long dark strings are Ti2O3 type reconstructions whereas the shorter, brighter and cross-linked rows are of the TiO2 type. The fine structure of the final state of the surface can be seen in Fig. 4 which shows the (1 × 1) structure of the ridge on the lower and upper CSP pair, the (1 × 2) strings forming from the upper CSPs. At the lower edge of the CSP pairs bright points can be seen terminating the Ti rows in the (1 × 1) ridge which are characteristic of under coordinated Ti ions. The lateral displacement of the (1 × 1) surface between the CSP pair and the surrounding (1 × 1) terrace is also evident.Fig. 5 shows a large area image of the surface after reoxidation with the same area marked for comparison to the initial state of the surface in Fig. 2. There has clearly been considerable growth with conversion of (1 × 2) terraces to (1 × 1) and the movement of normal step edges across the surface. The CSPs have, however, not moved in the surface [but have changed appearance due to the conversion of the terraces in which they are embedded from (1 × 2) to (1 × 1)]. The CSPs have grown in the <110> direction by 1 atomic layer. The lateral immobility of these 1/2 height steps in the surface is in stark contrast to the normal crystallographic steps which flow freely across the surface during oxidation.2 It is perhaps surprising that the CSPs remain visible on the surface during re-oxidation and are not buried by the regrowing layers of TiO2 as a result of the reaction between interstitial Tin+ and ambient O2. Furthermore while the CSPs act as nucleation centres for the growth of the TiO the process is apparently different for either side of the CSP pair at the surface. This observation is important for a number of reasons as it demonstrates the key role defect sites play in oxide surface chemistry and the influence of extended bulk defects on the surface structure. Furthermore it raises several questions, the most obvious being "why is there a difference in reactivity at the step edges of the CSP pair?" To address such a problem in detail would require a knowledge of the structure and mechanics of Tin+ interstitial diffusion, especially in the presence of CSPs, which is lacking.However, the STM gives us some important insights, enabling the discussion of possible mechanisms for the observed difference in reactivity. The first relates to the availability of Tin+ interstitials reaching the surface. The CSPs themselves may be expected to have a low permeability to diffusing interstitials in the bulk due to their atomic scale structure. The crystallographic displacement at the CSP effectively populates the interstitial site with the Ti4+ from the displaced lattice. In order to populate this site extra Ti is required during formation and is taken from the bulk interstitial concentration.The movement and formation of CSPs is kinetically hindered up to ~1000 K which suggests negligible diffusion of Tin+ through the CSP. Thus the available volume bounded by the surface and the CSP [and hence the number of interstitials free to re-oxidise (see Fig. 3b)] would differ on either side of a CSP which forms at an angle to the surface. This would naturally lead to a slower rate of growth on the acute step edge due to the lack of available Tin+. Second, the difference in structure at the two CSP 1/2 height step edges in the pair (one acute, one oblique) may lead to stabilisation of a preferred structure at one edge which nucleates further growth into the neighbouring terrace. In this case we would expect to observe different terminations with the surrounding terrace on either side of the CSP pair.In fact these are observable between the CSP pair and the original (1 × 2) surface as shown in Fig. 3a. In this example the surface was produced by annealing and should reflect the thermodynamically stable structures. Third, it is likely that the different step edges of the CSP pair may show differing activity to the dissociation of oxygen, which leads to the observable differences in growth rate. We have good evidence that such an effect would lead to an enhanced growth rate local to the active site. In a similar experiment in which Pd nanoparticles were deposited on the cross-linked (1 × 2) surface and then exposed to O2 at 673 K the surface reoxidised and grew faster in the area around the Pd.33,34 This is due to a much greater rate of oxygen adsorption and dissociation on the metal surface.The lateral extent of theenhanced growth region was temperature dependent. Therefore enhanced rates of oxygen dissociation at the CSP would be expected to lead to a locally higher growth rate. In all the above cases the surface diffusion of a mobile precursor to growth (such as TiO species) may play a key role. A precursor would lead to an increased ability to search for favourable growth sites, a reduced dependence on the local site of the initial oxygen adsorption and the ability to traverse CSPs on the surface. It should be noted that all these (and other) mechanisms may play a role to some degree in the re-oxidation.At present we have evidence for the formation of differing, thermodynamically stable structures at each step edge and for a locally higher rate of growth for a related system in which the supply of oxygen is localised. The observed growth of TiO2 on top of the central slab indicates that either a surface mobile precursor to growth exists or that the CSPs do not form efficient barriers to interstitial diffusion. Conclusions We have shown that small linear defect structures can be formed at the surface of cross-linked (1 × 2) reconstructed TiO2(110) and that these structures are crystallographic shear planes. The pairing of the CSPs at the surface leads to the formation of two half height step edges with locally different terminations to the surrounding terraces. These differing terminations result from the angle that the CSP makes to the surface leaving acute and obtuse step edges.On exposure to oxygen at elevated temperature the surface re-oxidises by removal of Tin+ interstitials dissolved in the crystal bulk to reform TiO2 on the surface. The CSPs nucleate the growth of this TiO2 with each step edge of the pair showing markedly different activity. This is manifested in a difference in growth rate and surface structures present at the step edges forming the pair of CSPs. The CSPs remain visible at the surface during re-oxidation. Acknowledgements The author would like to acknowledge the EPSRC for funding and would additionally like to thank Professor Michael Bowker for useful discussions and Peter Stone for experimental assistance.References 1 V. E. Henrich and P. A. Cox, The Surface Science of Metal Oxides, Cambridge University Press, Cambridge, 1996, p. 1. 2 G. S. Rohrer, V. E. Henrich and D. A. Bonnell, Science, 1990, 250, 1239. 3 H. Onishi and Y. Iwasawa, Phys. Rev. Lett., 1996, 76, 791. 4 H. Onishi and Y. Iwasawa, Surf. Sci., 1994, 313, L783. 5 D. Novak, E. Garfunkel and T. Gustafsson, Phys. Rev. B, 1994, 50, 5000. 6 M. Sander and T. Engel, Surf. Sci., 1994, 302, L263. 7 A. Szabo and T. Engel, Surf. Sci., 1995, 329, 241. 8 U. Diebold, J. F. Anderson, K. O. Ng and D. Vanderbilt, Phys. Rev. Lett., 1996, 77, 1322. 9 K. O. Ng and D. Vanderbilt, Phys. Rev.B, 1997, 56, 10544. 10 P. W. Murray, N. G. Condon and G. Thornton, Phys. Rev. B, 1995, 51, 10989. 11 C. L. Pang, S. A. Haycock, H. Raza, P. W. Murray, G. Thornton, O. Gülseren, R. James and D. W. Bullet, Phys. Rev. B, 1998, 58, 1586. 12 R. E. Tanner, M. R. Castell and G. A. D Briggs, Surf. Sci., 1998, 412–413, 672. 13 M. A. Henderson, Surf. Sci., 1999, 419, 174. 14 M. A. Henderson, Surf. Sci., 1995, 343, L1156. 15 W. S. Epling, C. H. F. Peden, M. A. Henderson and U. Diebold, Surf. Sci., 1998, 412–413, 333. 16 M. Li, W. Hebenstreit and U. Diebold, Surf. Sci., 1998, 414, L951. 17 P. Stone, R. A. Bennett and M. Bowker, New J. Phys., 1999, 1, 8, www.njp.org. 18 R. A. Bennett, P. Stone and M. Bowker, Phys. Rev. Lett., 1999, 82, 3831. 19 M. Li, W. Hebenstreit and U. Diebold, Phys. Rev. B, 2000, 61, 4926. 20 M. Li, W. Hebenstreit, L. Gross, U. Diebold, M. A. Henderson, D. R. Jennison, P. A. Schultz and M. P. Sears, Surf. Sci., 1999, 437, 173. 21 Q. Guo, I. Cocks and M. Williams, Phys. Rev. Lett., 1996, 77, 3851. 22 L. A. Bursill and B. G. Hyde, in Progress in Solid State Chemistry, ed. H. Reiss and J. O. McCaldin, Pergamon Press, New York, 1972, vol. 7, p. 177. 23 C. N. R. Rao and B. Raueau, Transition Metal Oxides, VCH, New York, 1995, p. 179. 24 M. G. Blanchin, L. A. Bursill and D. J. Smith, Proc. R. Soc. London, Ser. A, 1984, 391, 351. 25 J. B. Goodenough, Prog. Solid State Chem., 1972, 5, 145. 26 G. S. Rohrer, V. E. Henrich and D. A. Bonnell, Surf. Sci., 1992, 278, 146. 27 H. Nörenberg, R. E. Tanner, K. D. Schierbaum, S. Fischer and G. A. D. Briggs, Surf. Sci., 1998, 396, 52. 28 H. Nörenberg and G. A. D. Briggs, Surf. Sci., 1998, 402–404, 738. 29 R. A. Bennett, S. Poulston, P. Stone and M. Bowker, Phys. Rev. B, 1999, 59, 10341. 30 S. Bernal, F. J. Botana, J. J. Calvino, C. López, J. A. Pérez-Omil and J. M. Rodríguez-Izquierdo, J. Chem. Soc., Faraday Trans., 1996, 92, 2799. 31 B. Delmon, Surf. Rev. Lett., 1995, 2, 25. 32 M. Bowker, S. Poulston, R. A. Bennett, P. Stone, A. H. Jones, S. Haq and P. Hollins, J. Mol. Catal. A: Chem., 33 R. A. Bennett, P. Stone and M. Bowker, Catal. Lett., 1998, 131, 185. 1999, 59, 99. 34 R. A. Bennett, P. Stone and M. Bowker, Faraday Discuss., 1999, 114, 267. PhysChemComm © The Royal Society of Chemistry 2000

ISSN:1460-2733    

DOI:10.1039/b001938k

出版商:RSC    

年代:2000

数据来源: RSC